Getting Started

Welcome to the “Web-based Meta-Analysis Using R”. In this chapter, you can perform the first web-based meta-analysis of your own using the sample data in a minute.

Perform Meta-Analysis In A Minute

The first screen : Select language

When you load the meta-analysis app, you can see this screen. Currently, this app support English and Korean. You can select your preferred language(arrow).

1.png

Steps for meta-analysis

Performing meta-analysis with sample data is very simple. You can finish in five steps. Of these, just two clicks can do this.

  1. Select/Upload Data … use default data.
  2. Edit Data … No action required.
  3. Select Data Format … (1)
  4. Assign Variables …(arrow) No action required.

2.png

… And click the Meta Analysis button(2).

3.png

You can get the meta-analysis results as well as result of homogeneity test followed by the forest plot.

4.png

You can also get the radial plot(s) and the qqplot(s).

5.png

By default, you can also get the result of influential analysis and plots for influential diagnositics, DFBETAS plot followed by best linear unbiased prectictions(BLUP) and BLUP plot.

6.png 7.png

Download Results

You can download Report(1) as PDF(*.pdf), HTML(*.html,default) format. You can download plots(2) with png, svg or pdf format. You can download all results as a powerpoint file(*.pptx)(3). You can select editable vector graphics in powerpoint file(default).

8.png

Analyze Your Own Data

In this chapter, I will discuss how to upload your own data to the web-based meta-analysis app and edit the data. Also you will be able to learn how to use sample data as a template.

Upload data

You can upload your own data by clicking the choose file button(1). In the popup window, you can select your own data file. A data file with Microsoft excel format(*.xlsx) or comma-separate value format(*.csv) is allowed. Because an excel file contains calculations, functions or macros may cause error, the .csv format is preferred. You can save an excel file with a .csv format by “save as…” menu on excel. The limitation of file size is 30 MB. After selection of you data file, press choose button(arrow).

9.png

After a few seconds, the upload completed.

10.png

You should select the uploaded_file(1) among the Select Data radio buttons. Your file is displayed at Edit Data window.

11.png

Edit data

You can edit your data by click a cell in the table. You can use your data table as a spreadsheet.

12.png

Insert/remove row

By right click the table, you can insert row or remove row.

13.png

Use sample data as a template

You can use a sample data as a template. As a example, please choose the scared(binary)(1) among the Select Data radio buttons.

14.png

You can choose Remove all rows except 1(arrow) after right click of the table.

15.png

All rows are removed except the first row.

16.png

You are not allowed remove the first row, but you can edit the data. If you want to change the structure of your data, you should make a data file with the other program and upload it.

A Complete Example - Meta-analysis of BCG trials

In this chapter, I will show you how to perform subgroup analysis, meta-regression, publication bias analysis using funnel plot, cumulative meta-analysis, sensitivity test and download report as a pdf file or as a powerpoint file

Select Data

Please choose the bcg(binary2)(1) among the Select Data radio buttons. The data will be displayed(arrow).

17.png

Select Data Format

Please choose the Binary Data2 among the Select Data Format radio buttons. The variables will be assigned(arrow).

18.png

Select Additional Analysis

To perform subgroup analysis, select the subgroup analysis checkbox(1) and select alloc as a Grouping Variable(2). To perform a meta-analysis, select the Meta Regression checkbox(5) and select latitude as a Moderator(s)(4). To perform the publication bias test, cumulative meta-analysis and sensitivity test, select the checkboxes(5) and press the Meta-analysis button(arrow).

19.png

Customize Plot Options

You can customize your plot options by click the Show/hide Plot options(arrow).

20.png

You can choose whether or not all study labels be showed(1). You can label your studies by name or by study number(2. You can adjust the plot resolution and width and height of plot.

21.png

You can select some studies instead of labelling all studies.

23.png